The invention relates to adenoviruses. The large family of adenoviruses is subdivided according to its host into adenoviruses which infect mammals (the mastadenoviridae) and adenoviruses which infect birds (the aviadenoviridae). The CELO virus (Chicken Embryo Lethal Orphan; article by Mcferran, et al., 1977; McCracken and Adair, 1993) was identified as an infectious agent in 1957 (Yates and Fry, 1957). CELO virus is classified as a poultry adenovirus type 1 (FAV-1) and first aroused interest because of its property of being tumorigenic in baby hamsters. However, since infection with the CELO virus does not have any serious health and economic consequences, the interest in this virus disappeared in recent years. The FAV-1 adenoviruses can be isolated from healthy chickens and do not cause any disease when reintroduced experimentally into chickens (Cowen, et al., 1978). Their isolation from sick birds is presumably the result of adenovirus replication in a host which has an immune system weakened by other influences.
The general structural organisation of CELO virus, with an icosahedral capsid of 70-80 nm, made up of hexon and penton structures, is similar to that of the mammalian adenoviruses (Laver, et al., 1971). The CELO virus genome is a linear, double-stranded DNA molecule, the DNA being condensed inside the virion by virus-coded core proteins (Laver et al., 1971; Li, et al., 1984b). The CELO virus genome has covalently bound terminal proteins (Li, et al., 1983) and the genome has inverted terminal repeats (ITRs), although they are shorter than the mammalian ITRs (Alestrxc3x6m, et al., 1982b; Sheppard and Trist, 1992). The CELO virus codes a protease with 61-69% homology for the mammalian adenovirus proteases (Cai and Weber, 1993).
There are significant differences between CELO virus and the mastadenoviruses. CELO virus has a larger genome, with sequence homology with Ad5 which can only be detected in two short regions of the CELO virus genome (by hybridisation) (Alestrxc3x6m, et al., 1982a). The CELO virion has been reported to have two fibres of different lengths at each vertex. The CELO virus cannot complement the E1A functions of Ad5, and the replication of CELO virus is not made easier by the activity of Ad5E1 (Li, et al., 1984c).
Within the scope of the present invention, total sequence analysis of the CELO virus was carried out; on the one hand because it is useful for understanding the biology of adenoviruses to clarify the genomic organisation of an adenovirus which is very remote from the mammalian adenoviruses generally studied. Since the conditions for transmission and survival for a virus which infects a type of bird are presumably different than for mammalian viruses, it is possible that the bird adenoviruses have acquired new virus functions or exhibit a higher degree of variability than the mastadenoviridae. The complete CELO virus sequence also permits changes in the CELO virus genome with respect to functional analysis.
Since adenovirus vectors have proved highly effective vectors for gene transfer (see the summarising article by Graham, 1990; Kozarsky and Wilson, 1993; Trapnell and Gorziglia, 1994), the complete CELO virus sequence, on the other hands is particularly interesting as the basis for preparing new recombinant vectors for gene transfer.
Sequence analysis has shown that the CELO virus genome has 43.8 kb, being more than 8 kb longer than the human subtypes Ad2 and Ad5. The genes for the main structural proteins (hexon, penton based, IIIa, fibres, pVI, pVII, pVIII) are on the one hand both present and also located at the corresponding sites in the genome. The genes of the early region 2 (E2; DNA binding protein, DNA polymerase and terminal protein) are also present. However, the CELO virus lacks sequences homologous to the regions E1, E3 and E4 of the mammalian adenoviruses. There are approximately 5 kb at the left hand end and 15 kb at the right hand end of the CELO virus genome, where there is only restricted homology or no homology at all with the mastadenovirus genomes. These new sequences contain a number of open reading frames, and it can be assumed that these code for functions which replace the missing E1, E3 and possibly E4 regions.
Parts of the CELO virus sequence have already been published; they are listed in Table 1, as are the differences between the sequence known from the databank and the sequence determined within the scope of the present invention. From studies concentrating on specific viral genes, a homolog of the VA RNA gene of mastadenovirus was known (Larsson, et al., 1986) and part of the genome sequence which carries the endoprotease has been described (Cai and Weber, 1993). In addition, fragments of the CELO virus genome have been published (Akopian, et al., 1990; Akopian, et al., 1992; Hess, et al., 1995). The sequence of the penton base of the related virus FAV-10 has also been reported (Sheppard and Trist, 1992). Some other sequence fragments have been deposited in the databank and are also shown in Table 1. In all, about 50% of the CELO virus genome is available in the form of fragments (total about 24 kb). The sequence obtained within the scope of the present invention is complete and has the advantage of having been obtained from a single isolated material.
The total sequence of the CELO virus is shown in the sequence listing (in the sequence listing the word xe2x80x9ccomplementaryxe2x80x9d indicates that the open reading frames are present in the reverse arrangement). It shows a large number of striking differences between Ad2 and the CELO virus. The organisation of the recognisable open reading frames (ORFs) of the CELO virus genome based on the sequence analysis, compared with Ad2, is shown in FIG. 1A: the Figure shows an overview of the genomic organisation of Ad2/5 and CELO virus. The arrows indicate the position of the coding regions but not the exact cleavage patterns of the gene products. The pattern of the CELO virus also (in the first 6,000 bp and in the last 13,000 bp) gives all the non-associated open reading frames which begin with a methionine and which code for more than 99 amino acid groups. The central region of the two genomes which show homology on the basis of dot matrix analysis (cf. FIG. 3) and the regions at the ends of the CELO virus genome which have no homology with other adenoviruses (xe2x80x9cunique to CELOxe2x80x9d) are given. The abbreviations in the Figure, which also correspond to those in the Tables, have the following meanings: PB, penton base; EP, endoproteinase; DBP, DNA binding protein; bTP, preterminal protein; pol, DNA polymerase.
The sequenced CELO virus genome has a length of 43,804 bp and has a content of G+C of 54.3%. It had already been presumed at an earlier stage that the CELO virus genome is much larger than the mastadenovirus genome with 34-36 kb; it has been found that the CELO virus DNA has a weight of 30xc3x97106 Daltons, determined according to its sedimentation coefficient (Laver, et al., 1971), compared with 24xc3x97106 Daltons for Ad2 (Green, et al., 1967). The size of the CELO virus genome determined by the addition of the restriction fragments is about 43 kb (Cai and Weber, 1993; Denisova, et al., 1979). A Pulsed Field Gel Analysis of the CELO virus genome isolated from purified virions is shown in FIG. 2A and is compared with the DNA isolated from Ad5 dl1014 (34,600 bp; Bridge and Ketner, 1989) or Wild-type Ad5-virions (35,935 bp; vt300; Chroboczek, et al., 1992; Jones and Shenk, 1978); a mixture of uncleaved bacteriophage xcex-DNA and xcex3-DNA cleaved with five different restriction enzymes (Biorad) was used as the size marker (tracks 1 and 7 show the molecular weight markers, track 2 shows the DNA of Ad5 dl1014, track 3 shows the DNA of Ad5 wt300, track 4 shows the CELO virus DNA, track 5 shows the DNA of OTE, track 6 shows the DNA of Indiana C). FIG. 2A shows that the CELO virus genome has a length of 44 kb. From this analysis it is clear that the CELO virus genome is actually substantially larger than the genome of the mammalian virus. Calculations based on the migration of fragments of the lambda bacteriophage give a size of 43 kb for the CELO virus genome. The DNA extracted from two other FAV-1 isolates, Indiana C and OTE, co-migrates with the CELO virus species, which is further evidence of the size of the CELO virus genome. FIG. 2B shows that the CELO virus sequence contained in the bacterial plasmid pBR327 has the same size.
There is no identifiable E1 region. No significant homology could be found between the CELO virus genome and the first 4,000 bp of Ad2. There are a few small open reading frames in the first 5,000 bp of CELO virus which might possibly perform some of the E1 tasks. An open reading frame at the right hand end of the virus genome (GAM-1) may replace E1B 19K in functional assays without there being any significant homology between GAM-1 and E1B 19K. In order to confirm that the left hand end originates from the Wild type CELO virus genome and is not the sequence of a cloned variant, various tests were carried out; comparison of the direct sequence analysis of CELO virions at three different sites with the corresponding sites of the cloned sequences; Southern analyses with DNA from various virus isolates which yielded the same restriction fragments; pulsed field gel electrophoresis of various virus genomes which showed no heterogenicity.
There is no identifiable E3 region; the two small open reading frames in the corresponding region of the CELO virus have no significant homology with the E3 functions described.
There is a group of small open reading frames between 36,000 and 31,000 the position of which indicates the mammalian virus E4 region, but with additional 8 kb sequence at the right hand end of the CELO virus.
Nor was any sequence resembling protein IX identified (protein IX is essential for the hexon-hexon interactions and the stability of the mammalian adenovirus virions).
A protein V gene was not identified either.
The following regions are conserved between CELO virus and Ad2: the central part of the CELO virus genome, from the IVa2 gene (approximately from nucleotide (nt) 5,000) on the left hand strand up to the fibre genes on the right hand strand (approximately up to nt 33,000) is organised as in the mastadenoviruses, and the majority of the important viral genes can be identified both by their position and by sequence homology. Earlier studies on the homology between CELO and Ad2 (Alestrxc3x6m, et al., 1982a) showed two regions of the CELO virus which cross-hybridise with the Ad2 sequence. These two fragments are nt 5,626 to 8,877 (coding for IVa2 and the carboxy terminus of DNA polymerase) and nt 17,881 to 21,607 (coding for the hexon). The dot matrix analysis shown in FIG. 3 (carried out using the UWGCG program Compare with a window of 30 and a stringency of 20; summarised in FIG. 1A) shows that the total DNA sequence homology between CELO virus and Ad5 is mapped in the central region of the CELO virus genome. This is to be expected because the capsid proteins are coded in this central region and the coarse structure of the CELO virion is comparable with the capsid of the mammalian adenovirus (Laver, et al., 1971; Li, et al., 1984a). The genes which code for proteins corresponding to the human adenovirus proteins hexon, IIIa, penton base, protein VI and protein VIII, are present, and indeed in the expected sequence and position (FIG. 1A and Table 2A; Table 2B shows non-associated open reading frames which code for gene products with more than 99 amino acid groups). Each vertex of the mastadenovirus virion contains a pentamer of the penton base protein in conjunction with a single fibre consisting of three copies of the fibre polypeptide. Ad2, like most mastadenoviruses, has a single fibre gene, some adenovirus types have two fibre genes. The CELO virus genome codes for two fibre polypeptides of different lengths and sequences.
DNA binding proteins were identified in the region E2 (Li, et al., 1984c); four proteins with similar peptide maps were described, indicating a single precursor which is then cleaved or decomposed. The left hand open reading frame of the CELO virus genome, starting at nt 23,224, is located in the expected DNA binding protein region. The genes coding for DNA polymerase and pTP (pre-terminal protein) are present and in the expected positions (FIG. 1A, Table 2A).
With a view to the preparation of vectors based on the CELO virus it is useful to identify the mechanisms which the CELO virus uses in order to package almost 44 kb of DNA into a virion which is of a similar size to the human adenoviruses which are subjected to considerable restrictions on their packaging capacity (Bett, et al., 1993; Caravokyri and Leppard, 1995; Ghosh-Choudhury, 1987). One possibility is that the CELO virion, although virtually identical in size to Ad2 and Ad5, has a sufficiently widened structure in order to accommodate the larger genome. An alternative hypothesis is that CELO has a different mechanism for condensing DNA and therefore has differences in its provision of core proteins which are responsible for DNA packaging. Laver et al., in 1971 identified two proteins in the nucleus of the CELO virus and noted the absence of a molecule resembling protein V. Li, et al., 1984b, used electrophoresis with a higher resolution and reported a nuclear structure with three polypeptides (20 kD, 12 kD and 9.5 kD). These two findings lead one to conclude that the CELO virus must lack the larger basic nuclear protein V (41 kD) which occurs in mammalian adenoviruses. Perhaps the absence of protein V and/or the presence of smaller basic proteins is responsible for the additional packaging capacity of the CELO virion. The smallest of the CELO virus core proteins identified by Li, et al., 1984b (9.5 kD) is most closely associated with the virus DNA, similar to protein VII of the human adenovirus. An open reading frame which leads one to expect a protein with 8,597 D having 72 amino acids is located at nt 16,679; the coded protein is rich in arginine (32.9 mol %) and contains two cleavage sites for protease (pVII of Ad2 has only one cleavage site). An open reading frame which leads one to expect a protein with 19,777 D having 188 amino acid groups is located at nt 16,929. The protein has protease cleavage sites after groups 22, 128 and 145, and the carboxy-terminal groups have homology with pX of mastadenovirus. FIG. 4 shows the amino acid sequences of protein VII and pX of various mastadenoviruses compared with the CELO virus and the core proteins Core 2 and Core 1 of FAV-10. The sequences were arranged using the UWGCG Bestfit Program with a gap weight of 3.0 and a weight and gap length (xe2x80x9cGap Length Weightxe2x80x9d) of 0.1. The protease cleavage sites of adenovirus are underlined. In connection with this it is interesting that the mastadenovirus DNA binding protein designated xe2x80x9cmuxe2x80x9d consisting of 19 groups is formed by two protease cleavings of the pX precursor (Hosokava and Sung, 1976; Weber and Anderson, 1988; Anderson, et al., 1989). Cleaving of the protein having 188 groups after groups 128 and 145 would produce a mu-like basic protein consisting of 17 groups (41% arginine, 12% lysine). The uncleaved form of the protein is also highly basic; the uncleaved copies of this protein could correspond to the 20 kD core protein observed by Li, et al., 1984b; a third 12 kD core protein identified by these authors could not yet be assigned.
Moreover, some new or non-assigned open reading frames were found in the CELO virus genome. A summary of these open reading frames is shown in Table 2A; these open reading frames are also given in FIG. 1A. This summary was restricted to the sequences from nt 0-6,000 and 31,000-43,804 and only ORFs which contain a methionine group and code for a protein  greater than 99 amino acid groups are mentioned. As already stated, there is an ORF at nt 1999 which codes for a protein having homology with parvovirus-REP, and an ORF at nt 794 having homology with dUTPase and Ad2 E4 ORF1. The objective of the present invention was to prepare a new CELO virus.
Thus, on the basis of the complete CELO virus genome sequence, the present invention relates to a CELO virus obtained by in vitro manipulation of a plasmid-cloned CELO virus DNA.
The CELO virus according to the invention derived from the genomic DNA contains, in a preferred embodiment, the left and right terminal repeat and the packaging signal and has modifications in the form of insertions and/or deletions and/or mutations in regions of the CELO virus DNA which are different therefrom.
The left or right terminal repeat (xe2x80x9cInverted Terminal Repeatxe2x80x9d, ITR) extends from nucleotides 1-68 or from nucleotides 43734-43804, the packaging signal (also referred to as xe2x80x9cPsixe2x80x9d) extends from nucleotides 70-200. Modifications in DNA sections other than these ensures that the genes affected by the modification are non-functional or are deleted.
Preferably, modifications of the CELO virus genome are undertaken which are located on a section of the CELO virus DNA which includes the nucleotides from about 201 to about 5,000 (following the left terminal repeat of the section at the left hand end, hereinafter referred to as xe2x80x9cSection Axe2x80x9d) and/or on a section which includes the nucleotides from about 31,800- about 43,734 (the section at the right hand end located in front of the right terminal repeat, hereinafter referred to as xe2x80x9cSection Bxe2x80x9d) and/or on a section which includes the nucleotides from about 28,114-30,495 (the region of the fibre 1 gene, hereinafter referred to as xe2x80x9cSection Cxe2x80x9d).
A CELO virus in which certain genes are non-functional or are deleted, e.g. genes which affect the immune response of the host, such as antagonists to genes of the E3 region of mammalian adenoviruses, can be used as a vaccine.
In one embodiment of the invention the CELO virus contains one or more foreign DNA molecules, particularly a foreign DNA which is to be expressed in a host organism. In this embodiment the CELO virus acts as a vector which is capable of transporting the foreign DNA into higher eukaryotic cells, tissue or organisms, particularly mammals and birds, and expressing it therein.
Suitable insertion sites for the foreign DNA are the sections A and/or B and/or C.
The foreign DNA preferably replaces one or more sequences from these sections.
The CELO virus according to the invention is contained on a plasmid which is replicatable in bacteria or yeast and which yields virus particles after being introduced into suitable cells. Examples of suitable cells are bird embryo kidney or liver cells.
With a view to using a recombinant CELO virus vector for gene therapy, the foreign DNA may consist of one or more therapeutically active genes. Examples of these are genes coding for immunomodulators or modulators of inflammatory processes (cytokines such as IL-2, GM-CSF, IL-1, IL-6, IL-12; interferons, tumour antigens, IxcexaB, and derivatives of IxcexaB which lack serine phosphorylation sites (Traenckner, et al., 1995) or which lack lysine ubiquitinisation sites; glucocorticoid receptors; enzymes such as catalase, manganese superoxide dismutase, glutathione peroxidase, LIP members of the C/EBP family such as LIP or LAP (Descombes and Schibler, 1991), ADF (Tagaya, et al., 1989)), genes which influence apoptosis (members of the Bcl-2 family such as Bcl-2, adenovirus E1B19K, Mcl-2; BAX; IRF-2; members of the ICE protease family; variants of cJun, such as TAM-67 (Brown, et al., 1994); adenovirus E1A; p53) and genes which code for other therapeutic proteins (e.g. clotting factors such as factor VIII or IX; growth factors such as erythropoetin; cystic fibrosis transmembrane regulator gene (CFTR); dystrophin and derivatives thereof; globin; the LDL receptor; genes which are absent in lysosomal storage dysfunctions such as xcex2-glucuronidase; etc.).
With regard to the production of cellular tumour vaccines or for pharmaceutical compositions with which the immune response to tumours is to be intensified, the foreign DNA codes for immunostimulating proteins or tumour antigens or fragments thereof.
The therapeutically active DNA may also code for antisense molecules which prevent the expression of genes or the transcription of specific RNA sequences in the target cell.
With regard to the use of the recombinant CELO virus vector as a vaccine, the foreign DNA codes for one or more antigens which bring about an immune response in the individual treated.
In one embodiment of the invention the foreign DNA codes for an antigen derived from a human pathogen, particularly a pathogen of infectious diseases.
Epitopes which can be expressed by recombinant CELO viruses include epitopes derived from all kinds of human viral pathogens such as HIV, hepatitis A, B, C, hanta virus, polio virus, influenza virus, respiratory syncytial virus, measles, mumps, rubella, papilloma and many other viruses. The non-viral pathogens include trypanosomes (the causal agents of sleeping sickness and Chagas"" sickness), leishmania, Plasmodium falciparum (malaria), various bacterial pathogens such as the pathogens which cause tuberculosis, leprosy, Pseudomonas aeruginosa (complications in cystic fibrosis) and many others.
A summary of vaccines based on mastadenoviruses is given in Table 3; the epitopes mentioned therein by way of example can also be used for insertion in a vector based on the CELO virus.
With regard to the use of the recombinant CELO virus vector as a vaccine in the veterinary field, e.g. for birds, particularly poultry, the foreign DNA, in another preferred embodiment, codes for an antigen derived from a protein of a pathogen of animal diseases, particularly infectious diseases in birds.
Examples of pathogens of bird diseases are Avian Infectious Bronchitis Virus (IBV, a corona virus; Jia, et al., 1995; Ignjatovic and McWaters, 1991; Kusters, et al., 1990; Lenstra, et al., 1989; Cavanagh, et al., 1988; Cunningham, 1975), Avian Influenza Virus (Orthomyxovirus Type A; Kodihalli, et al., 1994; Treanor, et al., 1991; Tripathy and Schnitzlein, 1991), Fowlpox-Virus (McMillen, et al., 1994), Avian Infectious Laryngotracheitis Virus (Guo, et al., 1994; Scholz, et al., 1993; Keeler, et al., 1991), Mycoplasma Gallisepticum (Nascimento, et al., 1993), Avian Pasteurella Multocida (Wilson, et al., 1993; Lee, et al., 1991; Hertman, et al., 1980; Hertman, et al., 1979), Avian Reovirus (Ni and Kemp, 1992; Huang, et al., 1987), Marek""s Disease Virus (MDV; Malkinson, et al., 1992; Scott, et al., 1989), Herpes virus of turkeys (HVT, Herpes virus of Turkeys), Newcastle Disease Virus (NDV; Cosset, et al., 1991; Morrison, et al., 1990), Avian Paramyxovirus Type 1 (Jestin, et al., 1989), Avipoxvirus Isolates (Schnitzlein, et al., 1988), such as Juncopox, Pigeon Pox, and Field-(Field) and vaccine strains of bird pox viruses, Avian Encephalomyelitis Virus (Shafren and Tannock, 1991; Nicholas, et al., 1987; Deshmukh, et al., 1974), Avian Sarcoma Virus, Rotavirus (Estes and Graham, 1985), Avian Reovirus (Haffer, 1984; Gouvea, et al., 1983; Gouvea and Schnitzer, 1982), H7 Influenza Virus (Fynan, et al., 1993).
Apart from DNA sequences which code for therapeutically active gene products or for antigens, the foreign DNA may code for proteins or protein fragments which change the behaviour of the CELO virus, particularly its ability to bind to cells, with regard to the use of mammals, particularly on mammalian cells. Examples of such proteins are fibre or penton base proteins from mammalian and other adenoviruses, surface proteins of other viruses and ligands which have the ability to bind to mammalian cells and transport the CELO virus into the cells. Suitable ligands include transferrin from various mammalian species, lectins, antibodies or antibody fragments, etc. The skilled person will be aware of ligands of this kind, other examples can be found in WO 93/07283.
The recombinant CELO virus may contain one or more foreign DNA molecules. These may be inserted either in tandem or spaced apart in different sections of the CELO virus sequence.
The foreign DNA is under the control of regulatory sequences; suitable promoters include for example the CMV immediate early promoter/enhancer, the Rous sarcoma virus LTR, the adenovirus major late promoter and the CELO virus major late promoter.
The suitability of the CELO virus for preparing vectors and the advantages of these vectors and their applications are based particularly on the following properties of the CELO virus:
i) Safety: Naturally, the CELO virus does not replicate in mammalian cells. Therefore, vectors based on this virus can be used in humans without any danger of a subsequent infection with a Wild type human adenovirus complementing the vector and possibly allowing replication. This is an advantage over the Ad2 and Ad5 vectors used at present.
ii) Increased packaging capacity: The CELO virus genome is about 44 kb long, compared with the 36 kb of the Ad5 genome. Both viruses have comparable virion sizes, so that with a CELO virus vector it is possible to expand the strict packaging limit of 35 kb which is available with Ad5. On the basis of the sequencing carried out within the scope of the present invention, DNA-packaging core proteins of the CELO virus were identified, and striking differences were found from Ad2 which could be responsible for the increased packaging capacity. There are about 13 kb at both ends of the CELO virus which would appear not to code for structural proteins (e.g. capsid components) or for proteins required directly for the virus replication (e.g. DNA polymerase). For the production of vectors, these sequences on the CELO virus genome can be removed and if necessary replaced by complementing cell lines. These sequences can be assumed to code for the immune functions or the apoptotic functions (e.g. GAM-1) of the host cell or to be involved in the activation of the host cell for virus replication (antagonists to the E1, E3 and E4 regions of Ad2). These are the gene types which are either non-essential for virus growth in the cell culture, such as the E3 genes of Ad2 (Wold and Gooding, 1991; Gooding, 1992), or which are easily removed from the virus and can be expressed by a complementing cell line, such as the E1 region in 293 cells (Graham, et al., 1977) and the E4 region in W162 cells (Weinberg and Ketner, 1983).
iii) Stability: The CELO virion is remarkably stable. Its infectivity and its ability to transport DNA withstand a 30 minute treatment at 60xc2x0 C. As a comparison, Ad5 loses two powers of ten of its infectivity at 48xc2x0 C. and is completely inactivated at 52xc2x0 C. Presumably, the CELO virus did not develop its heat stability naturally, rather this heat stability would indicate a reaction to another type of selective pressure on the virion. The natural route of CELO virus infection is a faecal-oral route, which requires the virion to survive contact with a chemically aggressive environment with extreme pH values and with proteases. For particular applications in gene therapy, a more resistant virus would be desirable which would survive, for example, in the digestive tract or in the lungs of a patient suffering from cystic fibrosis.
iv) Targeted use: The CELO virus binds only slightly to mammalian cells on its own and for efficient entry into the cell requires the addition of a ligand (transferrin or lectin; Cotten, et al., 1993). Therefore, recombinant CELO virions are unable to penetrate into human cells, resulting in the following possible applications:
The virus may be genetically modified as stated above in order to express, on its surface, ligands which enable targeted transportation, such as for example specific peptides or fibres and/or penton bases of human adenoviruses.
Another possibility is the chemical modification of the virus in order to couple specific ligands such as transferrin thereto, as proposed for example in WO 94/24299 and additionally the virus may be biotinylated (WO 93/07283) and bound via streptavidin to biotinylated ligands such as wheatgerm agglutinin or other lectins (WO 93/07283). CELO virus vectors thus do not have the disadvantages of human adenoviruses which have good binding ability to human cells but have to be masked for a specific, targeted use mediated by the ligand.
v) Possibility of use for vaccines: The CELO virus is seldom connected to diseases in birds, indicating that it provokes a strong protective immune response in bird hosts. The CELO virus vector can easily be adapted for the expression of new vaccine epitopes.
With regard to the removal of regions of the CELO virus DNA, it was concluded from the results obtained with the mastadenovirus that, when the central sections of the CELO genome are removed, these sections have to be made available in trans, e.g. by a packaging cell line. This restriction is based on the large quantity of virion components which are necessary for the assembly of the virus and the need to produce a cell line which is capable of producing the corresponding amounts of these proteins without any toxicity.
The approaches which have hitherto proved more successful in this respect with mastadenoviruses consisted of the production of cell lines which express regulatory proteins (E1 and E4 regions) or enzymatic proteins (DNA polymerase, DNA binding protein), because these proteins are not required in large amounts during a productive virus infection.
Starting from the analysis of the known CELO genes, the sections of the genome from nt about 12,000 to about 33,000 which code for structural components of the virus are preferably not interrupted. The region from about nt 5,000 to about 12,000 codes for the E2 genes IVa2 (a viral transcription factor), viral DNA polymerase (POL) and the terminal virus protein (viral terminal protein; pTP) These genes are essential for the function of mastadenoviruses; they ought therefore to be essential for the CELO virus as well. However, it is theoretically also possible to have deletions of essential genes of this kind provided that they can be produced in trans, e.g. by a packaging line. For example, a packaging cell line was prepared which produces Ad5 DNA polymerase, thus making it possible for this gene to be deleted from the virus genome. A similar procedure may be used in the construction of CELO virus vectors, by removing sections or the entire region from nt 5,000 to 12,000 from the CELO virus and having the corresponding functions controlled in trans by a packaging cell line.
Another possible restriction exists with regard to the presumed major late promoter, which was provisionally assigned to the region at about nt 7,000 (TATA box at nt 7,488). In the mastadenoviruses, this promoter is essential for driving late gene expression. Therefore, any change in the region at nt 7,000 of the CELO genome must be carried out in a way which maintains the promoter function of this region.
Table 4 lists the sequence elements of the CELO virus genome and is divided into various categories with regard to their deletion and/or mutation in the production of CELO virus vectors (in Table 4 L1, L2, etc. denote xe2x80x9clate message 1, 2, etc., corresponding to the nomenclature normally used for mastadenoviruses):
Category 1 includes sequence elements which are required in cis and therefore cannot be made available in trans by a complementing cell line or by a complementing plasmid. These sections are necessary and are therefore present on the CELO virus according to the invention; these are the left and right terminal repeats and the packaging signal.
Sequences of category 2 code for proteins which are required in large amounts for virion production. These proteins may optionally be produced by a gene contained in a complementing cell line or on a complementing plasmid. Other sequences of category 2 are the major later promoter, the tripartite leader sequence and also the splice acceptor sites (SA) or the polyadenylation sites (poly A sites) of genes which are essential and which cannot be made available in trans.
Fundamentally, care must be taken to ensure that, in modifications of the CELO virus DNA which are carried out at the boundaries of genes, any control signals present, e.g. polyA sites are not interrupted or affected in any other way if possible.
The genes deleted from the CELO virus or non-functional genes therein may be prepared in trans, e.g. by complementing cell lines.
Complementing cell lines (xe2x80x9chelper cellsxe2x80x9d) can be prepared by the manner known from the literature, analogously to helper cells which complement the functions of mammalian adenoviruses. To do this, the relevant CELO virus gene on a plasmid, preferably combined with a selectable marker, is introduced into cells which permit the replication of CELO virus, preferably into immortalised cell lines such as LMH (Kawaguchi, et al., 1987) or immortalised quail cell lines, as described for example by Guilhot et al., 1993. In helper cells which express the relevant CELO virus genes, optionally stably integrated, the defective CELO viruses are able to replicate.
Instead of making the regions of the CELO virus deleted in the vector available by means of a cell line, the deletions may also be complemented by a copy of the relevant gene contained on a plasmid. For this, the method described in WO 96/03517, that described by Cotten et al., 1994a and 1994b) or the one described by Wagner et al., 1992, may be used, for example, in which a CELO virus vector containing a deletion is inserted, as a component of a transfection complex, containing a conjugate of polylysine and a UV/psoralen-inactivated adenovirus (human or CELO) and optionally transferrin-polylysine, into embryonic chicken kidney cells or liver cells, embryonic or immortalised quail cells, e.g. liver or kidney cells, and the transfection complex also contains a plasmid which carries a copy of the gene or genes which lack the CELO virus vector. The combination of genes contained on the vector and genes carried by the plasmid results in a normal virus replication cycle. (Similar approaches were used in mastadenovirus systems in order to complement E1-deficient adenoviruses (Goldsmith, et al., 1994) and E4-deficient adenoviruses (Scaria, et al., 1995)). The subsequent amplification of the virus may be carried out by using the defective virus as a carrier which is dependent on the complementing plasmid using the methods described above.
Another possible way of replacing the genes missing from the CELO virus is by using helper viruses.
The helper virus used may be a CELO virus (Wild type or partly defective). In this embodiment the CELO plasmid carrying the mutation (e.g. a derivative of pCELO7), is introduced into chicken cells, e.g. using the method described in WO 96/03517 or described by Cotten et al., 1994, using as the carrier for the derivative, for example, psoralen/UV-inactivated adenovirus (human or CELO) together with an adenovirus (human or CELO) as a carrier for the plasmid or plasmids with the genes which complement the defect. Alternatively, a Wild type CELO virus may be used both as a carrier and as source for complementing gene functions. The subsequent amplification of the defective CELO viruses obtained is carried out by co-infection of the defective CELO virus with a complementing adenovirus (e.g. Wild-type CELO or a CELO which has mutations at other points of the genome).
CELO virus genes of category 3 include the sequences on Sections A, B and C. These are sequences which code for a protein or an RNA molecule which is necessary for the interaction with the host cell machinery or with the host immune system. These proteins should be required in fairly low concentrations or may be dispensable for cultivation of the virus in the tissue culture.
Thus, preferably, in the CELO virus vectors according to the invention the genes of category 3 are replaced by the gene in question; if necessary, complementary cell lines or plasmids or helper viruses can be prepared which produce the corresponding gene products.
In one embodiment of the invention the vectors according to the invention contain the gene in question instead of one of the fibre genes. The CELO virus has two fibre proteins (Laver, et al., 1971; Gelderblom and Maichle-Lauppe, 1982; Li, et al., 1984a). It can be assumed that one of the fibres of the CELO virus is not necessary for the assembly of the virion and the infectivity. This assumption is backed up by electron microscopic observations that the longer fibre (fibre 1) should associate with the penton base along the side of the complex, whilst the shorter fibre (fibre 2) projects out of the middle of the penton base, similarly to the penton/fibre complexes in the mastadenoviruses (Hess, et al., 1995). In adenoviruses with only a single fibre, the fibre molecule is required for the assembly of the virus; in the absence of fibres no stable mature viruses are formed. The CELO virion should therefore require fibre 2 for stability and as a ligand, whereas fibre 1 acts only as a ligand. Within the scope of the present invention the assumption that, of the two fibre genes of the CELO virus, fibre gene 1 located in region C is superfluous and can be replaced by the gene in question was proved correct by removing the fibre 1 gene and replacing it with a luciferase expression unit.
Other-examples are inserts in region A and/or B.
Within the scope of the present invention it was found that destruction of the reading frame at nt 794 (region A) which codes for dUTPase yields viable viruses. The dUTPase gene is thus a gene which is not necessary for growth in cell culture.
In one embodiment of the invention, the recombinant CELO virus thus contains a foreign gene which is inserted in the region of the reading frame coding for dUTPase.
According to another aspect the present invention relates to a process for preparing recombinant CELO virus.
The process is characterised in that the CELO virus genome contained on a plasmid or sections thereof is or are genetically manipulated.
In one aspect of the invention the genetic manipulation consists of insertion and/or deletion. Insertions and/or deletions can be carried out by using restriction enzyme cutting sites which occur naturally in the CELO virus DNA in these sections, e.g. the FseI cutting site occurring at position 35,693 in Section B. The insertion may be carried out directly into this cutting site or beyond this cutting site or close to this cutting site, or this cutting site may be used to allow recombination in the surrounding area.
In one preferred embodiment the manipulation consists of carrying out insertions and/or deletions using standard methods of molecular biology (Maniatis, 1989). The naturally occurring restriction cutting sites can be used for this, e.g. sites located in regions of the genome which are non-essential for cultivation of the virus in the host cell, e.g. the FseI cutting site which occurs at position 35,693 in Section B. The insertion can be made directly into this cutting site or beyond this cutting site or in the vicinity of this cutting site, or the cutting site may be used to facilitate recombination in the surrounding area. Foreign DNA sequences can be inserted, e.g. marker genes or genes coding for therapeutically active proteins.
An alternative possibility is to remove CELO sequences which are flanked by two restriction sites and replace them with new sequences. (An example of this is dUTPase mutation as carried out in Example 7.) In these cases the manipulation is carried out with the entire CELO virus genome. In the event that restriction sites are present, the deletion/insertion can alternatively be carried out on a subfragment which is then re-incorporated in the entire genome by ligation and optionally recloning in bacteria.
Another possibility is to insert the foreign gene in artificial restriction enzyme cutting sites produced by conventional methods of recombinant DNA technology (Maniatis, 1989).
In one embodiment the process is characterised in that manipulations are carried out in a plasmid DNA which contains the CELO virus genome, in CELO DNA sequences, with the exception of the left and right inverted terminal repeats and the packaging signal.
In another preferred method, the manipulation of the CELO genome is carried out by recombination. For this, a subfragment of the CELO genome is manipulated in order to introduce mutations and/or new sequences. Subfragments can be prepared by various methods, specifically by PCR (polymerase chain reaction), by ligation between PCR products or between restriction fragments or by subcloning in bacteria (as described in the Examples of the previous invention; see also Chartier et al., 1996). Examples of suitable bacteria strains for recombination are BJ 5183 (Hanahan, 1983) or JC 8679 (Gillen et al., 1974) or JC 5176 (Capado-Kimball and Barbour, 1971).
For recombination using PCR products, the sequence to be inserted into the CELO genome is prepared by PCR (Oliner et al., 1993) using primers which flank the sequence plus about 15 nucleotides of the sequence complementary to the insertion site in the CELO genome. In a second round of PCR, another 15 nucleotides are hung from the sequence complementary to CELO, resulting in a PCR product which consists of the sequence to be inserted with 30 nucleotides of the CELO sequence at each end. This fragment is mixed with a plasmid which contains the CELO DNA (e.g. the plasmid pCELO7 prepared within the scope of the present invention) and which has been linearised with a restriction enzyme which cuts only between the two flanking sequences hung from the sequence by PCR.
For recombination using ligation reaction products (prepared by conventional techniques as described for example by Maniatis et al., 1989), in principle the same procedure is used as in the recombination with cloned fragments, except that the intermediate cloning step is omitted.
In every case, the manipulated product obtained is characterised and used to prepare virus by transfecting avian cells (e.g. using the method described by Wagner et al., 1992; Cotten et al., 1994; or Cotten et al., 1993) and then cultivating them, after which the virus is harvested.
For preparing recombinant CELO virus using cloned fragments, the method preferably comprises subcloning a small fragment from the relevant region of CELO virus into which the foreign gene is to be inserted on a bacterial plasmid in order to ensure that restriction sites which occur several times on the CELO virus genome occur only once on the plasmid. These restriction sites are used to remove a region from the small fragment. For preparing the CELO virus vector this region is replaced by foreign DNA. The foreign DNA may consist solely of a linker with a restriction site occurring only once or of a sequence coding for a protein or for an antigen. The sequence may also code for a reporter gene with a restriction site occurring only once. This makes further manipulation of the CELO virus easier by inserting the foreign DNA, which codes for a therapeutically active gene product or for an antigen, into this restriction site, and at the same time the reporter gene permits rapid information as to the efficiency of the vector, by introducing the plasmid into cells and monitoring the expression of the reporter gene.